Classification of Aurora‐A Kinase Inhibitors Using Self‐Organizing Map (SOM) and Support Vector Machine (SVM)
Classification of Aurora‐A Kinase Inhibitors Using Self‐Organizing Map (SOM) and Support Vector Machine (SVM)
复制标题
DOI：
10.1002/minf.201000106
复制
发表时间：
2011-01
影响因子：
3.6
通讯作者：
Liyu Wang;Zhi Wang;A. Yan;Qipeng Yuan
中科院分区：
文献类型：
--
作者：
Liyu Wang;Zhi Wang;A. Yan;Qipeng Yuan
建立了148种极光激酶抑制剂的两种分类模型，以区分极光激酶的活性和弱效活性抑制剂。每个分子由ADRIANA.Code计算的12个选定的分子描述符表示。然后分别利用Kohonen自组织图（SOM）和支持向量机（SVM）方法建立分类模型，对现有数据库进行虚拟筛选，寻找可能具有较高活性的新先导化合物。SOM模型对训练集和测试集的预测准确率分别为96.6%和90.0%，SVM模型对训练集和测试集的预测准确率分别为93.2%和93.3%。
Two classification models of 148 Aurora‐A kinase inhibitors were developed to separate active and weakly potent active inhibitors of Aurora‐A kinase. Each molecule was represented by 12 selected molecular descriptors calculated by the ADRIANA.Code. Then the classification models were built using a Kohonen’s Self‐Organizing Map (SOM) and a Support Vector Machine (SVM) method, respectively, which could be used for virtual screening an existing database to find possible new lead compounds with higher activity. The prediction accuracy of the models for the training and test sets are 96.6 % and 90.0 % for SOM, 93.2 % and 93.3 % for SVM.